Basic OSS concepts

Last Updated: Sep 25, 2017

This section introduces the basic concepts of OSS.


An object (also known as a file) is a discrete unit of data.

An object is composed of:

  • Metadata, known as Object Meta, which is a key-value pair that expresses the object’s attributes, such as its last modification time and size, as well as user-defined information.
  • User data, known as Data.
  • A unique object name, known as a Key.

The size of an object varies with the upload method. Multipart Upload supports objects of up to 48.8 TB. Other upload methods only support objects of up to 5 GB.

An object’s lifecycle starts from when it has been successfully uploaded, and ends when it has been deleted. During an object’s lifecycle, its information cannot be changed. If you upload an object with a duplicate name in a bucket, it will overwrite the existing one. Therefore, unlike the file system, OSS does not allow users to modify only part of an object.

OSS provides the Append Upload function, which allows users to continually append data to the end of an object.

The name of an object must comply with the following rules:

  • It must use UTF-8 encoding.
  • It must be between 1-1023 bytes in length.
  • It cannot start with a backslash “/“ or forward slash “\”.

NOTE: Object names are case sensitive.


A bucket is a virtual division of object storage that, unlike file systems, manages objects in a flat structure.

Bucket properties are as follows:

  • All objects must belong to a bucket, and during an object’s lifecycle it remains directly affiliated with the corresponding bucket.
  • You can have multiple buckets, with each bucket able to contain an unlimited number of objects.
  • You can set and modify the attributes of a bucket for region and object access control and object lifecycle management. These attributes apply to all objects in the bucket.
  • You can create different buckets to perform different management functions.

The name of a bucket must comply with the following rules:

  • It can only contain lower-case letters, digits, and hyphens (-).
  • It must start and end with a lower-case letter or number.
  • It must be between 3-63 bytes in length.
  • It must be globally unique within the OSS.

Once a bucket name is created, it cannot be changed.


A region represents the physical location of an OSS data center.

Users can select regions based on fees, request sources, and other factors according to site requirements. Generally, the closer the user is in proximity to a region, the faster the access speed will be. For more information, see OSS Regions and Endpoints.

A region is specified when a bucket is created, and cannot be changed. All objects contained in this bucket are therefore stored in the corresponding data center. Setting different regions for objects in the same bucket is currently not supported.


An endpoint is the domain name used to access the OSS.

OSS provides external services through HTTP RESTful APIs. Different regions use different endpoints. For the same region, access through an intranet or through the Internet also uses different endpoints. For example, regarding the Hangzhou region:

  • The intranet endpoint is
  • The Internet endpoint is

For more information, see OSS Regions and Endpoints.


An AccessKey (AK) is composed of an AccessKeyId and AccessKeySecret. The AccessKeyId is a public key, whereas the AccessKeySecret is a private key and must be kept confidential. These are then paired together to perform access identity verification.

The OSS verifies the identity of a request sender by using the AccessKeyId/AccessKeySecret symmetric encryption method. The AccessKeyId identifies a user. With the AccessKeySecret, a user can encrypt the signature string. The OSS then uses the symmetric encryption method to verify the AccessKey of the signature string. In OSS, AccessKeys are generated as follows:

  • Applied for by the bucket owner.
  • Granted by the bucket owner to an authorized third-party requestor through RAM.
  • Granted by the bucket owner to an authorized third-party requestor through STS.

For more information about AccessKeys, see RAM.

High consistency

In OSS, object operations are binary, that is, operations must either succeed or fail without an intermediate status. After a user uploads an object, OSS ensures that it is complete. OSS will not return a partial success response when uploading objects.

Object operations in OSS are likewise highly consistent. For example, once a user receives an upload (PUT) success response, this object can be read immediately, and the data will have already been written in triplicate. The same concept applies to delete operations. Once a user deletes an object, this object no longer exists.

This high-consistency feature facilitates user architectural design. The logic of OSS usage is the same as that of a traditional storage device: modifications are immediately visible and users do not have to consider final consistency issues.

Comparison between OSS and file system

OSS is a distributed object storage service structure that uses a Key-Value pair format, whereas a file system uses a tree-type index structure of directories that contain files. In OSS, users retrieve object content based on unique object names (Keys). In file systems, users retrieve files based on their location in a directory.

The benefit of OSS is that it supports massive concurrent access volumes, which means large volumes of unstructured data (such as images, videos, and documents) can be stored and retrieved without excessive use of resources. The benefit of a file system is that folder operations such as renaming, moving, and deleting directories (which means renaming, moving, and deleting data) is considerably easier as data does not need to be copied and replaced.

The limitation of OSS is that saved objects cannot be modified. If an object needs modification, the entire object must be uploaded again to make the modification take effect. One exception is through using the append object operation, whereby users call a specific API, which allows a generated object be of a different type than normally uploaded objects. The limitations of a file system are that system performance is limited to a single device, and the more files and directories that are created in the system, the more resources are consumed, and the lengthier user processes become.

Comparisons between OSS and file system concepts are as follows:

OSS File system
Object File
Bucket Main directory
Region N/A
Endpoint N/A
AccessKey N/A
N/A Multilevel directory
GetService Retrieving the list of main directories
GetBucket Retrieving the list of files
PutObject Writing an object
AppendObject Append writing an object
GetObject Reading an object
DeleteObject Deleting an object
N/A Modifying file content
CopyObject (same target and source) Modifying file attributes
CopyObject Copying an object
N/A Renaming an object

OSS Glossary

Term Definition
Object A discrete unit of data
Bucket A virtual division of object storage
Endpoint The domain name for OSS access
Region A representation of the physical location of an OSS data center
AccessKey An alias for the AccessKeyId and AccessKeySecret pair
Put Object Simple upload
Post Object Form upload
Multipart Upload The uploading of an object as several chunks, then reassembling the chunks
Append Object An upload that attaches to already uploaded data
Get Object Simple download
Callback Upload callback
Object Meta Metadata of a file that includes the object’s attributes and user-defined information
Data Object information (typically user-defined information)
Key Unique object name
ACL (Access Control List) Permissions for buckets or files
Thank you! We've received your feedback.